Staying Safe: Spotting AI Scams & Deepfakes

AI Makes Scams More Convincing

Scammers now use AI to create fake voices, realistic emails, and convincing deepfake videos. The good news? Once you know what to look for, AI scams are still spottable.

The New AI Scam Toolkit

What scammers can do with AI now:

• Clone voices from short audio samples (10 seconds)
• Write personalized phishing emails at scale
• Generate fake profile pictures that look real
• Create deepfake videos of people saying things they never said
• Automate romance scams with chatbots
• Produce fake websites and documents instantly

But they still make mistakes. This lesson teaches you to spot them.

Voice Cloning Scams (The Emergency Call)

How it works:

1. Scammer finds voice sample (social media video, voicemail, public speaking)
2. AI clones the voice in minutes
3. Calls family member pretending to be you in "emergency"
4. Asks for immediate money transfer

Real example:

"Mom, I'm in trouble. I crashed the car and need $5,000 for bail. Can you send it right now? Don't tell Dad, I'm scared."

Red flags:

• Urgent request for money
• Asks you not to tell anyone else
• Unusual payment method (gift cards, crypto, wire transfer)
• Won't video call or answer specific questions
• Background noise sounds artificial
• Emotional manipulation ("I'm scared," "Please help")

What to do:

1. Hang up immediately
2. Call the person directly on a number you already have saved
3. Ask a personal question only they would know
4. Never send money without verifying in person or via video call
5. Set up a family code word for real emergencies

Family safety tip:
Agree on a code word with family members. In a real emergency, they'll use it. Scammers won't know it.

Deepfake Videos: How to Spot Them

Deepfakes are AI-generated videos that make people appear to say or do things they never did.

Common uses:

• Fake celebrity endorsements
• Political misinformation
• Impersonating CEOs or public figures
• Creating fake "proof" of events

How to spot deepfakes:

Look at the face:

• Unnatural blinking (too little or too much)
• Weird shadows on face
• Face doesn't match neck/body skin tone
• Teeth look blurry or unnatural
• Hair has strange artifacts or fuzziness

Look at the edges:

• Face outline looks pasted on
• Edges of face have weird halos or blur
• Ears or hairline look inconsistent

Listen to audio:

• Lip sync is slightly off
• Voice sounds robotic or flat in emotion
• Background noise cuts out unnaturally
• Breathing doesn't match speaking pattern

Check context:

• Is this video surprising or out of character?
• Is it only on one sketchy website?
• Can you find the original source?
• Are reputable news sources reporting it?

Quick verification:

1. Reverse image search a frame from the video
2. Check if the person's verified social media confirms it
3. Look for fact-checks on Snopes, AP, Reuters
4. Ask: "Would this person really say this?"

AI-Generated Phishing Emails

Old phishing emails: Obvious typos, broken English, generic greetings
AI-powered phishing: Perfect grammar, personalized, sounds professional

What AI changes:

• No more "Dear Sir/Madam"—now it's your actual name
• No typos or weird phrasing
• Researches you on LinkedIn to sound credible
• Mimics writing style of people you know

Red flags that AI can't hide:

Urgency and threats:

"Your account will be closed in 24 hours!"
"Immediate action required!"

Unusual requests:

"Verify your account by clicking this link"
"Update payment information here"
"Confirm this wire transfer"

Sender email is slightly off:

• paypa1.com instead of paypal.com
• support@amaz0n.com instead of amazon.com
• Random numbers/letters in domain

Requests for sensitive info:

• Passwords
• Social Security number
• Credit card details
• Bank account info

What to do:

1. Don't click links in suspicious emails
2. Go directly to the website by typing the URL yourself
3. Call the company using the number from their official website
4. Hover over links (don't click) to see real destination
5. Forward to spam and delete

Fake AI-Powered Services

Scammers advertise "AI services" that don't exist or don't work.

Common scams:

• "AI lawyer" (generates invalid legal documents)
• "AI tax filing" (steals your data)
• "AI investment advisor" (loses your money)
• "AI resume builder" (charges for free tools)

Red flags:

• Suspiciously low prices
• No real company information
• Can't verify credentials (lawyer license, certification)
• Only accepts payment upfront
• No refund policy
• Advertised only on social media

How to verify:

1. Search "[Company name] + scam" on Google
2. Check Better Business Bureau (BBB)
3. For lawyers: Verify license with state bar
4. For financial advisors: Check SEC or FINRA
5. Read reviews on independent sites (not their website)

Romance Scams with AI Chatbots

AI catfishing is creating fake dating profiles with AI-generated photos and chatbot conversations.

How it works:

1. AI generates attractive profile photos (person doesn't exist)
2. AI chatbot manages conversation (can talk to hundreds at once)
3. Builds emotional connection over weeks/months
4. Eventually asks for money ("emergency," "travel to visit you")

Red flags:

• Profile photos look like professional models
• Professes love very quickly
• Refuses to video call (always has excuse)
• Has dramatic life circumstances or sob story
• Eventually asks for money
• Wants to move off dating app immediately
• Can't meet in person (always has excuse)

Protection:

• Reverse image search profile photos (right-click → Search Google for image)
• Insist on video call early in conversation
• Never send money to someone you haven't met in person
• Tell friends about new connections (they can spot red flags)
• Trust your gut if something feels off

Fake Reviews and AI-Generated Content

AI can write thousands of fake reviews in minutes.

Spotting fake reviews:

• Generic language: "This product exceeded my expectations!"
• Posted in short time span (50 reviews in one day)
• All 5-star or all 1-star (no variety)
• Reviews don't mention specific features
• Overly emotional or dramatic
• Same writing style across different "reviewers"

Tools to detect fake reviews:

• Fakespot (Fakespot.com)
• ReviewMeta (ReviewMeta.com)
• Both analyze patterns and flag suspicious reviews

Protecting Older Family Members

Seniors are often targeted because they're less familiar with AI scams.

How to help:

Set up defenses:

• Enable spam filters on email and phone
• Add caller ID apps (shows "Likely Spam")
• Set up two-factor authentication on important accounts
• Install ad blocker to reduce scam ads

Teach one rule:

"If anyone asks for money urgently, hang up and call me first."

Practice together:

• Show them examples of phishing emails
• Demonstrate how to verify suspicious calls
• Set up family code word for emergencies
• Practice saying "I need to verify this" and hanging up

Stay involved:

• Ask about strange calls or emails
• Review accounts together occasionally
• Don't shame them if they're uncertain—celebrate checking first

What to Do If You've Been Scammed

Act immediately:

If you sent money:

1. Contact your bank/credit card company right away (may reverse charge)
2. File police report
3. Report to FBI IC3 (ic3.gov)
4. Report to FTC (reportfraud.ftc.gov)
5. If gift cards: Contact the gift card company immediately
6. If wire transfer: Contact transfer service (Western Union, MoneyGram)

If you shared personal information:

1. Change passwords on all important accounts
2. Enable two-factor authentication everywhere
3. Place fraud alert on credit reports (call one bureau, they notify others)
4. Monitor bank and credit card statements daily
5. Consider credit freeze to prevent new accounts being opened
6. File identity theft report at IdentityTheft.gov

If it was a dating/romance scam:

1. Cut all contact with the scammer
2. Report to dating platform
3. Don't blame yourself—these are professional criminals
4. Seek support from friends or counselor

Report it:

• Reporting helps authorities track and stop scammers
• Helps protect others from the same scam
• May help you recover funds in some cases

AI Safety Checklist for Everyday Use

For emails:

• Sender's email address looks legitimate (check carefully)
• No urgent demands for action
• Not asking for passwords or financial info
• Links hover-preview shows correct website
• You were expecting this email

For phone calls:

• You recognize the number OR can verify caller
• No urgent requests for money/gift cards
• Caller can answer personal questions
• You can call back on official number
• It's okay to hang up and verify independently

For videos/images:

• Source is reputable (news org, verified account)
• Face movements look natural
• Audio syncs with lips
• You can find confirmation elsewhere
• Context makes sense

For online services:

• Company has verifiable credentials
• Reviews on independent sites are positive
• Prices are reasonable (not suspiciously cheap)
• Clear refund policy
• Secure payment (credit card, PayPal—not gift cards)

Practical Exercises

Exercise 1: Test your family's defenses
Send a family member a suspicious-looking email (that you created) and see if they spot it. Discuss what made it obvious.

Exercise 2: Reverse image search practice
Find a random profile photo online. Practice reverse image searching it on Google Images. See if you can find where it came from.

Exercise 3: Set up family code word
Choose a family emergency code word. Share it with parents, kids, siblings. Test it once to make sure everyone remembers.

Exercise 4: Spot the deepfake
Search "deepfake examples" on YouTube. Practice spotting the tells (face edges, blinking, lip sync).

Exercise 5: Enable two-factor authentication
Set up 2FA on your three most important accounts (email, bank, social media). Use an authenticator app (Google Authenticator, Authy) instead of SMS if possible.

Remember: When in Doubt, Verify

The golden rule:

If something feels urgent, unusual, or too good to be true—STOP and verify independently.

It's always okay to:

• Hang up and call back on a verified number
• Ignore an email and go directly to the website
• Say "I need to verify this first"
• Ask a friend or family member "Does this seem legit?"
• Take your time instead of acting immediately

Scammers rely on urgency. Remove urgency, and you remove their power.